search

conceptofmind / LaMDA-rlhf-pytorch

Open-source pre-training implementation of Google's LaMDA in PyTorch. Adding RLHF similar to ChatGPT.
MIT License
459 stars 76 forks source link

Hi, when i try to run it, it shows the error below #3

Open HoxtonR opened 2 years ago

HoxtonR commented 2 years ago

Traceback (most recent call last): File "C:\Users\HoxtonR\Desktop\LaMDA\LaMDA-pytorch-main\test.py", line 1, in from lamda_pytorch.lamda_pytorch import LaMDA File "C:\Users\HoxtonR\Desktop\LaMDA\LaMDA-pytorch-main\lamda_pytorch__init__.py", line 1, in from lamda_pytorch.lamda_pytorch import LaMDA File "C:\Users\HoxtonR\Desktop\LaMDA\LaMDA-pytorch-main\lamda_pytorch\lamda_pytorch.py", line 9, in from config.config import CFG ModuleNotFoundError: No module named 'config'

conceptofmind commented 2 years ago

Hi @HoxtonR ,

I just pushed an update to fix the import errors you had listed. One of the last updates had changed the file directory structure.

The repository initially consisted of just the pre-training architecture for the model but I am actively updating it on a daily basis to include a full pre-training example. The training script is not fully implemented as a tokenizer needs to still be provided and the package manager needs to be implemented since this model requires specific libraries for scaling.

I will be updating consistently and keep this issue open for now.

Best,

Enrico